Three studies are described. The first study is a test of two hypotheses, that there is a nonverbal "communication skill deficit" among dissatisfied marriages and a "private message system" that enhances communication among satisfied marriages. The second study is a 2x2 factorial random sample study (blue/white collar and satisfied/dissatisfied) of interaction at home and in the laboratory, the couple's personal history and their philosophies of marriage. The third study is a study of emotion and its role in marital satisfaction; it combines interaction data with physiological data and self-report of affect in a videotape recall session.